Experimenting OAR in a virtual cluster environment for batch schedulers comparative evaluation

Size: px
Start display at page:

Download "Experimenting OAR in a virtual cluster environment for batch schedulers comparative evaluation"

Transcription

1 Experimenting OAR in a virtual cluster environment for batch schedulers comparative evaluation Joseph Emeras (Joseph.Emeras@imag.fr) Yiannis Georgiou (Yiannis.Georgiou@imag.fr) 1 Context OAR [3] is the Grid 5000 Resources and Jobs Management System (RJMS). RJMS are composed of two parts: a batch scheduler and a resources manager. Their role is to plan and manage jobs executions on the resources and to manage the resources themselves. OAR is a tool for both production or research clusters. Its mechanism principle is the following: a user requests some resources associated to a task. The system schedules this request depending on availabilities and returns either a date when the task will start or a shell on one of the attributed resources. The OAR software is developed by MESCAL and MOAIS teams in the LIG [1] laboratory. Developments are also supported by the Aladdin/Grid 5000 community and the CIMENT [2] project, that are currently its two majors users. In addition, other installations were also listed in Brazil, Slovaquia, China and in France, where the BRGM in Orleans uses it. Currently, OAR handles succesfully up to several hundreds of resources on each site but we know that its scheduler may not be as efficient with thousands or hundreds of thousands resources. Since the number of cores and cpus per machine is increasing considerably, we must look into the future and be prepared to face the question of large scale resource management that could involve deep modifications in the data structures representing manipulated resources and scheduling algorithms. This is why we want to study: How OAR behaves comparing to the other RJMS depending on the number of resources through a comparative valuation How OAR will handle large scale resources An other unconfessed goal of such experiments in a virtual context is to ease OAR testing and tuning by helping the development team to test the new features they just coded and to set up non-regression tests in a real cluster context and researchers to test their new ideas and to play with OAR s tuning parameters. But since this is not the main goal, we will not range over too much this subject in this article. 1

2 2 Experiment goals We ve seen in the previous section that we want to study two aspects. For the comparative evaluation between OAR and other RJMS, several series of benchmarks have been made on several RJMS thanks to Grid 5000 and the benchmark suite ESP [6]. Amongst them, were OAR, Torque/MAUI and SLURM. The work was around the comparison between these systems depending on the number of managed resources. The problem here, is that such tests need a lot of resources on the Grid 5000 platform and it is not always easy to obtain the resources if a lot of jobs are running, and certainly not nice to the other users to monopolize the platform. The second aspect joins here the first one, if we want to study how OAR will handle large scale resources, we will need such large scale resources, and for the moment, this is not the case in Grid In fact, what we want to do, is to increase the number of resources in Grid 5000 to serve our goals. One response to this problem is Virtualization. If we manage to create several virtual machines per node on our cluster, we have solved our two problems. The next figure presents the concept of the virtualization in the experiment. In fact, if we manage to create a tool that will build automatically an OAR powered cluster inside Grid 5000, and if this tool is convenient enough to tune parameters such as the number of real nodes, the number of virtual machines per node, the virtual machines memory and the number of virtual cpus (vcpus) on the virtual machines, we will be able to play with the benchmarking tool and make very complete behaviour graphs of the different RJMS but also with few resources, we will be able to create a lot of virtual resources, that will be very usefull to test OAR in a large scale resources context. An other benefit of Virtualization will be to speed up the number of experiments results by allowing to run more experiments in parallel or to save energy by reducing the number of physical machines used for a same amount of experiments. It still remains a problem that we have to solve: will the use of virtual machines be transparent for benchmarking, in other words, if the use of virtualization slows down too much the system, the benchmark can be distorded. We have to take into account this problem in our choices. 2

3 Figure 1: Virtualization experiment concept. 3

4 3 Tools used We saw that we will need a lot of virtual machines, and that the global overhead due to virtualization has to be as small as possible. An other aspect here, is that by using many virtual machines on one host, this will induce a large amount of disk space usage. If one virtual machine takes about 400MiB, having dozens not to say hundreds of them could be problematic. To solve these two problems we chose to use Xen + LVM. Two studies of such problematics ([4] and [5]) shown that Xen is a good, reliable, and efficient virtualization tool. Furthermore, in several Grid 5000 sites, Xen powered images are available or will be soon. LVM and snapshot system will allow to minimize the disk space usage. By using this method, only the base image of the guests OS will take about 400MiB but the snapshots themselves will take only few KiB. At the end, having one virtual machine or hundreds of virtual machines on the same host will use approximately the same disk space. The benchmark tool used will be ESP (Effective System Performance). It is designed for RJMS testing. It launches jobs of different durations that will take from one resource to the entire cluster. These jobs can be (depending on how was parametrized the ESP run) simple sleeps or sort of cpuburn jobs. Here, we do not focus on the jobs themselves, their result is useless in a benchmark case, the only important thing is their run time, which is a parameter of the job itself and do not depend on the cpu speed nor on the current cpu charge. This is why in our case, it is not a problem to use virtualization and to run X jobs on one real core (X beeing the number of virtual machines on the host), because these jobs will still last the same duration as if there where only one job on the core. Since the goal of the experiment is to launch a lot of benchmark tests on different situations and then, study the results depending on the number of real machines, virtual machines, memory and vcpus allocated per vm, the process have to be automated. Thus, the Kapharnaum tool has been created. Currently it is more a prototype then a real production tool but it is reliable enough for experimenting the automated process. It takes as parameter these variables and then automatically creates the OAR powered virtual cluster on top of Xen and LVM snapshots. Everything is automatic, from the OAR server set-up and virtual computing nodes register to the benchmark launching and the results collect. It is to be said that when creating the nodes and server, the script installs OAR from the latest generated packages, this allowing, coupled to a smart automated package generation tool, to test the latest code commits. This tool, very easy to handle and mostly highly customizable will allow us to study deeply OAR s behaviour. But first, let s see if Xen handles well virtualization without a big overhead. 4

5 4 Results Now that we have a tool to test OAR in a virtualized context, it is important to compare its effectiveness in a regular vs. virtualized environment. By regular, we mean in a standard Grid 5000 cluster context. For now, we don t have a lot of results since the virtual cluster creation tool is very new, but some of them are very interesting. They are shown in the table below. Comparison of OAR ranks in virtualized vs. non-virtualized modes depending on the number of resources and the experiment parameters. Griffon cluster, non-virtualized mode Total resources number Number of real nodes 8 64 Number of cores per node 8 8 ESP rank (percent) Grelon cluster, virtualized mode Total resources number Number of real nodes Number of cores per node Virtual machines per node cores/vcpus used ESP rank (percent) Variance Griffon cluster, virtualized mode Total resources number Number of real nodes Number of cores per node Virtual machines per node cores/vcpus used ESP rank (percent) Variance Experiments realised in the Nancy Grid 5000 site on the griffon cluster (nodes: 92, cpuarch: x86-64, cputype: xeon-hapertown, nodecpu: 2, cpucore: 4, memnode: 16384, disktype: sata, ethnb: 1) and grelon cluster (nodes: 120, cpuarch: x86-64, cputype: xeon-woodcrest, nodecpu: 2, cpucore: 2, memnode: 2048, disktype: sata, ethnb: 2). The Total resources number is the Number of real nodes Virtual machines per node cores/vcpus used. It is important to say that the Kapharnaum tool is really new and is still in testing state, so we don t have enough results to make a valuable comparison. However, regarding the preliminary results we can see a tendency here. A benchmark with ESP will give comparable results if the number of total vcpus per host (number of vm on the host number of vcpus per vm) is less or equal 5

6 to eight time the number of real cores on the physical node. If this total number is bigger, we begin to observe some performance lost. Thus we can multiply by 8 the number of resources with XEN without impacting our benchmark result. These preliminary results encourage us to continue the work done here and to add more smartness and error control in Kapharnaum in order to observe more finely the XEN behaviour in this case. In conclusion, we can say that the Kapharnaum tool is almost ready for automated OAR benchmarks runs but it is still a work in progress if we want more precise knowledge of the interactions between XEN, ESP and OAR and if we want to improve the results. 6

7 References [1] Grenoble Computing Laboratory - [2] The ciment project - [3] N. Capit. A batch scheduler with high level components. CCGrid 05. [4] L. Nussbaum. Contributions à l expérimentation sur les systèmes distribués de grande taille - [5] B. Quetier. Etude de mécanismes de virtualisation pour l émulation conforme de Grilles à grande échelle. [6] A. T. Wong. ESP: A System Utilization Benchmark. 7

Provisioning and Resource Management at Large Scale (Kadeploy and OAR)

Provisioning and Resource Management at Large Scale (Kadeploy and OAR) Provisioning and Resource Management at Large Scale (Kadeploy and OAR) Olivier Richard Laboratoire d Informatique de Grenoble (LIG) Projet INRIA Mescal 31 octobre 2007 Olivier Richard ( Laboratoire d Informatique

More information

Enabling Large-Scale Testing of IaaS Cloud Platforms on the Grid 5000 Testbed

Enabling Large-Scale Testing of IaaS Cloud Platforms on the Grid 5000 Testbed Enabling Large-Scale Testing of IaaS Cloud Platforms on the Grid 5000 Testbed Sébastien Badia, Alexandra Carpen-Amarie, Adrien Lèbre, Lucas Nussbaum Grid 5000 S. Badia, A. Carpen-Amarie, A. Lèbre, L. Nussbaum

More information

PES. Batch virtualization and Cloud computing. Part 1: Batch virtualization. Batch virtualization and Cloud computing

PES. Batch virtualization and Cloud computing. Part 1: Batch virtualization. Batch virtualization and Cloud computing Batch virtualization and Cloud computing Batch virtualization and Cloud computing Part 1: Batch virtualization Tony Cass, Sebastien Goasguen, Belmiro Moreira, Ewan Roche, Ulrich Schwickerath, Romain Wartel

More information

U-LITE Network Infrastructure

U-LITE Network Infrastructure U-LITE: a proposal for scientific computing at LNGS S. Parlati, P. Spinnato, S. Stalio LNGS 13 Sep. 2011 20 years of Scientific Computing at LNGS Early 90s: highly centralized structure based on VMS cluster

More information

HPC performance applications on Virtual Clusters

HPC performance applications on Virtual Clusters Panagiotis Kritikakos EPCC, School of Physics & Astronomy, University of Edinburgh, Scotland - UK pkritika@epcc.ed.ac.uk 4 th IC-SCCE, Athens 7 th July 2010 This work investigates the performance of (Java)

More information

Computing in High- Energy-Physics: How Virtualization meets the Grid

Computing in High- Energy-Physics: How Virtualization meets the Grid Computing in High- Energy-Physics: How Virtualization meets the Grid Yves Kemp Institut für Experimentelle Kernphysik Universität Karlsruhe Yves Kemp Barcelona, 10/23/2006 Outline: Problems encountered

More information

Basics of Virtualisation

Basics of Virtualisation Basics of Virtualisation Volker Büge Institut für Experimentelle Kernphysik Universität Karlsruhe Die Kooperation von The x86 Architecture Why do we need virtualisation? x86 based operating systems are

More information

How do Users and Processes interact with the Operating System? Services for Processes. OS Structure with Services. Services for the OS Itself

How do Users and Processes interact with the Operating System? Services for Processes. OS Structure with Services. Services for the OS Itself How do Users and Processes interact with the Operating System? Users interact indirectly through a collection of system programs that make up the operating system interface. The interface could be: A GUI,

More information

High-Availability Using Open Source Software

High-Availability Using Open Source Software High-Availability Using Open Source Software Luka Perkov Iskon Internet, Zagreb, Croatia Nikola Pavković Ruđer Bošković Institute Bijenička cesta Zagreb, Croatia Juraj Petrović Faculty of Electrical Engineering

More information

Flauncher and DVMS Deploying and Scheduling Thousands of Virtual Machines on Hundreds of Nodes Distributed Geographically

Flauncher and DVMS Deploying and Scheduling Thousands of Virtual Machines on Hundreds of Nodes Distributed Geographically Flauncher and Deploying and Scheduling Thousands of Virtual Machines on Hundreds of Nodes Distributed Geographically Daniel Balouek, Adrien Lèbre, Flavien Quesnel To cite this version: Daniel Balouek,

More information

Improving Job Scheduling by using Machine Learning & Yet an another SLURM simulator. David Glesser, Yiannis Georgiou (BULL) Denis Trystram(LIG)

Improving Job Scheduling by using Machine Learning & Yet an another SLURM simulator. David Glesser, Yiannis Georgiou (BULL) Denis Trystram(LIG) Improving Job Scheduling by using Machine Learning & Yet an another SLURM simulator David Glesser, Yiannis Georgiou (BULL) Denis Trystram(LIG) Improving Job Scheduling by using Machine Learning & Yet an

More information

Best Practices for Monitoring Databases on VMware. Dean Richards Senior DBA, Confio Software

Best Practices for Monitoring Databases on VMware. Dean Richards Senior DBA, Confio Software Best Practices for Monitoring Databases on VMware Dean Richards Senior DBA, Confio Software 1 Who Am I? 20+ Years in Oracle & SQL Server DBA and Developer Worked for Oracle Consulting Specialize in Performance

More information

Takahiro Hirofuchi, Hidemoto Nakada, Satoshi Itoh, and Satoshi Sekiguchi

Takahiro Hirofuchi, Hidemoto Nakada, Satoshi Itoh, and Satoshi Sekiguchi Takahiro Hirofuchi, Hidemoto Nakada, Satoshi Itoh, and Satoshi Sekiguchi National Institute of Advanced Industrial Science and Technology (AIST), Japan VTDC2011, Jun. 8 th, 2011 1 Outline What is dynamic

More information

PARALLELS CLOUD SERVER

PARALLELS CLOUD SERVER PARALLELS CLOUD SERVER Performance and Scalability 1 Table of Contents Executive Summary... Error! Bookmark not defined. LAMP Stack Performance Evaluation... Error! Bookmark not defined. Background...

More information

UPS battery remote monitoring system in cloud computing

UPS battery remote monitoring system in cloud computing , pp.11-15 http://dx.doi.org/10.14257/astl.2014.53.03 UPS battery remote monitoring system in cloud computing Shiwei Li, Haiying Wang, Qi Fan School of Automation, Harbin University of Science and Technology

More information

PARALLELS CLOUD STORAGE

PARALLELS CLOUD STORAGE PARALLELS CLOUD STORAGE Performance Benchmark Results 1 Table of Contents Executive Summary... Error! Bookmark not defined. Architecture Overview... 3 Key Features... 5 No Special Hardware Requirements...

More information

Management of a Grid Infrastructure in glite with Virtualization

Management of a Grid Infrastructure in glite with Virtualization Management of a Grid Infrastructure in glite with Virtualization Miguel Cardenas Montes Javier Perez-Griffo Callejon Raul Ramos Pollan Manuel Rubio del Solar Ibergrid 2007 Santiago Compostela / 16 of May

More information

Virtualization with Windows

Virtualization with Windows Virtualization with Windows at CERN Juraj Sucik, Emmanuel Ormancey Internet Services Group Agenda Current status of IT-IS group virtualization service Server Self Service New virtualization features in

More information

A quantitative comparison between xen and kvm

A quantitative comparison between xen and kvm Home Search Collections Journals About Contact us My IOPscience A quantitative comparison between xen and kvm This content has been downloaded from IOPscience. Please scroll down to see the full text.

More information

Table of Contents. P a g e 2

Table of Contents. P a g e 2 Solution Guide Balancing Graphics Performance, User Density & Concurrency with NVIDIA GRID Virtual GPU Technology (vgpu ) for Autodesk AutoCAD Power Users V1.0 P a g e 2 Table of Contents The GRID vgpu

More information

STeP-IN SUMMIT 2013. June 18 21, 2013 at Bangalore, INDIA. Performance Testing of an IAAS Cloud Software (A CloudStack Use Case)

STeP-IN SUMMIT 2013. June 18 21, 2013 at Bangalore, INDIA. Performance Testing of an IAAS Cloud Software (A CloudStack Use Case) 10 th International Conference on Software Testing June 18 21, 2013 at Bangalore, INDIA by Sowmya Krishnan, Senior Software QA Engineer, Citrix Copyright: STeP-IN Forum and Quality Solutions for Information

More information

Load Balancing in the Cloud Computing Using Virtual Machine Migration: A Review

Load Balancing in the Cloud Computing Using Virtual Machine Migration: A Review Load Balancing in the Cloud Computing Using Virtual Machine Migration: A Review 1 Rukman Palta, 2 Rubal Jeet 1,2 Indo Global College Of Engineering, Abhipur, Punjab Technical University, jalandhar,india

More information

ZEN LOAD BALANCER EE v3.02 DATASHEET The Load Balancing made easy

ZEN LOAD BALANCER EE v3.02 DATASHEET The Load Balancing made easy ZEN LOAD BALANCER EE v3.02 DATASHEET The Load Balancing made easy OVERVIEW The global communication and the continuous growth of services provided through the Internet or local infrastructure require to

More information

Operating Systems Virtualization mechanisms

Operating Systems Virtualization mechanisms Operating Systems Virtualization mechanisms René Serral-Gracià Xavier Martorell-Bofill 1 1 Universitat Politècnica de Catalunya (UPC) May 26, 2014 Contents 1 Introduction 2 Hardware Virtualization mechanisms

More information

7/15/2011. Monitoring and Managing VDI. Monitoring a VDI Deployment. Veeam Monitor. Veeam Monitor

7/15/2011. Monitoring and Managing VDI. Monitoring a VDI Deployment. Veeam Monitor. Veeam Monitor Monitoring a VDI Deployment Monitoring and Managing VDI with Veeam Aseem Anwar S.E. Channel UKI Need for real-time performance metrics Detailed alerting and fault finding tools Identification of bottlenecks

More information

Intro to Virtualization

Intro to Virtualization Cloud@Ceid Seminars Intro to Virtualization Christos Alexakos Computer Engineer, MSc, PhD C. Sysadmin at Pattern Recognition Lab 1 st Seminar 19/3/2014 Contents What is virtualization How it works Hypervisor

More information

Monitoring Databases on VMware

Monitoring Databases on VMware Monitoring Databases on VMware Ensure Optimum Performance with the Correct Metrics By Dean Richards, Manager, Sales Engineering Confio Software 4772 Walnut Street, Suite 100 Boulder, CO 80301 www.confio.com

More information

WebLogic on Oracle Database Appliance: Combining High Availability and Simplicity

WebLogic on Oracle Database Appliance: Combining High Availability and Simplicity WebLogic on Oracle Database Appliance: Combining High Availability and Simplicity Frances Zhao-Perez Alexandra Huff Oracle CAF Product Management Simon Haslam Technical Director O-box Safe Harbor Statement

More information

IOS110. Virtualization 5/27/2014 1

IOS110. Virtualization 5/27/2014 1 IOS110 Virtualization 5/27/2014 1 Agenda What is Virtualization? Types of Virtualization. Advantages and Disadvantages. Virtualization software Hyper V What is Virtualization? Virtualization Refers to

More information

Ressources management and runtime environments in the exascale computing era

Ressources management and runtime environments in the exascale computing era Ressources management and runtime environments in the exascale computing era Guillaume Huard MOAIS and MESCAL INRIA Projects CNRS LIG Laboratory Grenoble University, France Guillaume Huard MOAIS and MESCAL

More information

Cloud Computing PES. (and virtualization at CERN) Cloud Computing. GridKa School 2011, Karlsruhe. Disclaimer: largely personal view of things

Cloud Computing PES. (and virtualization at CERN) Cloud Computing. GridKa School 2011, Karlsruhe. Disclaimer: largely personal view of things PES Cloud Computing Cloud Computing (and virtualization at CERN) Ulrich Schwickerath et al With special thanks to the many contributors to this presentation! GridKa School 2011, Karlsruhe CERN IT Department

More information

Toad for Oracle 8.6 SQL Tuning

Toad for Oracle 8.6 SQL Tuning Quick User Guide for Toad for Oracle 8.6 SQL Tuning SQL Tuning Version 6.1.1 SQL Tuning definitively solves SQL bottlenecks through a unique methodology that scans code, without executing programs, to

More information

Virtual Machine Synchronization for High Availability Clusters

Virtual Machine Synchronization for High Availability Clusters Virtual Machine Synchronization for High Availability Clusters Yoshiaki Tamura, Koji Sato, Seiji Kihara, Satoshi Moriai NTT Cyber Space Labs. 2007/4/17 Consolidating servers using VM Internet services

More information

Energy Efficient MapReduce

Energy Efficient MapReduce Energy Efficient MapReduce Motivation: Energy consumption is an important aspect of datacenters efficiency, the total power consumption in the united states has doubled from 2000 to 2005, representing

More information

ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy

ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy OVERVIEW The global communication and the continuous growth of services provided through the Internet or local infrastructure require to

More information

BackupEnabler: Virtually effortless backups for VMware Environments

BackupEnabler: Virtually effortless backups for VMware Environments White Paper BackupEnabler: Virtually effortless backups for VMware Environments Contents Abstract... 3 Why Standard Backup Processes Don t Work with Virtual Servers... 3 Agent-Based File-Level and Image-Level

More information

Review from last time. CS 537 Lecture 3 OS Structure. OS structure. What you should learn from this lecture

Review from last time. CS 537 Lecture 3 OS Structure. OS structure. What you should learn from this lecture Review from last time CS 537 Lecture 3 OS Structure What HW structures are used by the OS? What is a system call? Michael Swift Remzi Arpaci-Dussea, Michael Swift 1 Remzi Arpaci-Dussea, Michael Swift 2

More information

HyperV_Mon 3.0. Hyper-V Overhead. Introduction. A Free tool from TMurgent Technologies. Version 3.0

HyperV_Mon 3.0. Hyper-V Overhead. Introduction. A Free tool from TMurgent Technologies. Version 3.0 HyperV_Mon 3.0 A Free tool from TMurgent Technologies Version 3.0 Introduction HyperV_Mon is a GUI tool for viewing CPU performance of a system running Hyper-V from Microsoft. Virtualization adds a layer

More information

The QEMU/KVM Hypervisor

The QEMU/KVM Hypervisor The /KVM Hypervisor Understanding what's powering your virtual machine Dr. David Alan Gilbert dgilbert@redhat.com 2015-10-14 Topics Hypervisors and where /KVM sits Components of a virtual machine KVM Devices:

More information

Evaluation of Nagios for Real-time Cloud Virtual Machine Monitoring

Evaluation of Nagios for Real-time Cloud Virtual Machine Monitoring University of Victoria Faculty of Engineering Fall 2009 Work Term Report Evaluation of Nagios for Real-time Cloud Virtual Machine Monitoring Department of Physics University of Victoria Victoria, BC Michael

More information

JProfiler: Code Coverage Analysis Tool for OMP Project

JProfiler: Code Coverage Analysis Tool for OMP Project CMU 17-654 & 17-754 Analysis of Software Artifacts Spring 2006 Individual Project: Tool Analysis May 18, 2006 Eun-young Cho echo1@andrew.cmu.edu JProfiler: Code Coverage Analysis Tool for OMP Project Table

More information

VMware Server 2.0 Essentials. Virtualization Deployment and Management

VMware Server 2.0 Essentials. Virtualization Deployment and Management VMware Server 2.0 Essentials Virtualization Deployment and Management . This PDF is provided for personal use only. Unauthorized use, reproduction and/or distribution strictly prohibited. All rights reserved.

More information

Virtualization. Types of Interfaces

Virtualization. Types of Interfaces Virtualization Virtualization: extend or replace an existing interface to mimic the behavior of another system. Introduced in 1970s: run legacy software on newer mainframe hardware Handle platform diversity

More information

Virtual Machine Monitors. Dr. Marc E. Fiuczynski Research Scholar Princeton University

Virtual Machine Monitors. Dr. Marc E. Fiuczynski Research Scholar Princeton University Virtual Machine Monitors Dr. Marc E. Fiuczynski Research Scholar Princeton University Introduction Have been around since 1960 s on mainframes used for multitasking Good example VM/370 Have resurfaced

More information

The Top Six Advantages of CUDA-Ready Clusters. Ian Lumb Bright Evangelist

The Top Six Advantages of CUDA-Ready Clusters. Ian Lumb Bright Evangelist The Top Six Advantages of CUDA-Ready Clusters Ian Lumb Bright Evangelist GTC Express Webinar January 21, 2015 We scientists are time-constrained, said Dr. Yamanaka. Our priority is our research, not managing

More information

Red Hat Enterprise Virtualization - KVM-based infrastructure services at BNL

Red Hat Enterprise Virtualization - KVM-based infrastructure services at BNL Red Hat Enterprise Virtualization - KVM-based infrastructure services at Presented at NLIT, June 16, 2011 Vail, Colorado David Cortijo Brookhaven National Laboratory dcortijo@bnl.gov Notice: This presentation

More information

COS 318: Operating Systems. Virtual Machine Monitors

COS 318: Operating Systems. Virtual Machine Monitors COS 318: Operating Systems Virtual Machine Monitors Kai Li and Andy Bavier Computer Science Department Princeton University http://www.cs.princeton.edu/courses/archive/fall13/cos318/ Introduction u Have

More information

my forecasted needs. The constraint of asymmetrical processing was offset two ways. The first was by configuring the SAN and all hosts to utilize

my forecasted needs. The constraint of asymmetrical processing was offset two ways. The first was by configuring the SAN and all hosts to utilize 1) Disk performance When factoring in disk performance, one of the larger impacts on a VM is determined by the type of disk you opt to use for your VMs in Hyper-v manager/scvmm such as fixed vs dynamic.

More information

High Availability for Virtualized Environment. NEC Corporation

High Availability for Virtualized Environment. NEC Corporation High Availability for Virtualized Environment http://www.nec.com/expresscluster NEC Corporation System Software Division December, 2012 Ensuring High Availability In Virtual Environment In the virtual

More information

PARALLELS SERVER BARE METAL 5.0 README

PARALLELS SERVER BARE METAL 5.0 README PARALLELS SERVER BARE METAL 5.0 README 1999-2011 Parallels Holdings, Ltd. and its affiliates. All rights reserved. This document provides the first-priority information on the Parallels Server Bare Metal

More information

Container-based operating system virtualization: a scalable, high-performance alternative to hypervisors

Container-based operating system virtualization: a scalable, high-performance alternative to hypervisors Container-based operating system virtualization: a scalable, high-performance alternative to hypervisors Soltesz, et al (Princeton/Linux-VServer), Eurosys07 Context: Operating System Structure/Organization

More information

An Experimental Study of Load Balancing of OpenNebula Open-Source Cloud Computing Platform

An Experimental Study of Load Balancing of OpenNebula Open-Source Cloud Computing Platform An Experimental Study of Load Balancing of OpenNebula Open-Source Cloud Computing Platform A B M Moniruzzaman 1, Kawser Wazed Nafi 2, Prof. Syed Akhter Hossain 1 and Prof. M. M. A. Hashem 1 Department

More information

Dynamic resource management for energy saving in the cloud computing environment

Dynamic resource management for energy saving in the cloud computing environment Dynamic resource management for energy saving in the cloud computing environment Liang-Teh Lee, Kang-Yuan Liu, and Hui-Yang Huang Department of Computer Science and Engineering, Tatung University, Taiwan

More information

Characterizing Task Usage Shapes in Google s Compute Clusters

Characterizing Task Usage Shapes in Google s Compute Clusters Characterizing Task Usage Shapes in Google s Compute Clusters Qi Zhang 1, Joseph L. Hellerstein 2, Raouf Boutaba 1 1 University of Waterloo, 2 Google Inc. Introduction Cloud computing is becoming a key

More information

Mark Bennett. Search and the Virtual Machine

Mark Bennett. Search and the Virtual Machine Mark Bennett Search and the Virtual Machine Agenda Intro / Business Drivers What to do with Search + Virtual What Makes Search Fast (or Slow!) Virtual Platforms Test Results Trends / Wrap Up / Q & A Business

More information

Optimizing Shared Resource Contention in HPC Clusters

Optimizing Shared Resource Contention in HPC Clusters Optimizing Shared Resource Contention in HPC Clusters Sergey Blagodurov Simon Fraser University Alexandra Fedorova Simon Fraser University Abstract Contention for shared resources in HPC clusters occurs

More information

Maximizing SQL Server Virtualization Performance

Maximizing SQL Server Virtualization Performance Maximizing SQL Server Virtualization Performance Michael Otey Senior Technical Director Windows IT Pro SQL Server Pro 1 What this presentation covers Host configuration guidelines CPU, RAM, networking

More information

PARALLELS CLOUD SERVER

PARALLELS CLOUD SERVER PARALLELS CLOUD SERVER An Introduction to Operating System Virtualization and Parallels Cloud Server 1 Table of Contents Introduction... 3 Hardware Virtualization... 3 Operating System Virtualization...

More information

Grid vs. Cloud Computing

Grid vs. Cloud Computing Grid vs. Cloud Computing The similarities and differences between Cloud Computing and Extreme-Scale Computation on Demand 2008 Parabon Inc. All rights reserved. 2009 Parabon 1 Computation, Inc. All rights

More information

Virtualization Technologies and Blackboard: The Future of Blackboard Software on Multi-Core Technologies

Virtualization Technologies and Blackboard: The Future of Blackboard Software on Multi-Core Technologies Virtualization Technologies and Blackboard: The Future of Blackboard Software on Multi-Core Technologies Kurt Klemperer, Principal System Performance Engineer kklemperer@blackboard.com Agenda Session Length:

More information

Violin: A Framework for Extensible Block-level Storage

Violin: A Framework for Extensible Block-level Storage Violin: A Framework for Extensible Block-level Storage Michail Flouris Dept. of Computer Science, University of Toronto, Canada flouris@cs.toronto.edu Angelos Bilas ICS-FORTH & University of Crete, Greece

More information

Manjrasoft Market Oriented Cloud Computing Platform

Manjrasoft Market Oriented Cloud Computing Platform Manjrasoft Market Oriented Cloud Computing Platform Innovative Solutions for 3D Rendering Aneka is a market oriented Cloud development and management platform with rapid application development and workload

More information

Overlapping Data Transfer With Application Execution on Clusters

Overlapping Data Transfer With Application Execution on Clusters Overlapping Data Transfer With Application Execution on Clusters Karen L. Reid and Michael Stumm reid@cs.toronto.edu stumm@eecg.toronto.edu Department of Computer Science Department of Electrical and Computer

More information

Vmware Training. Introduction

Vmware Training. Introduction Vmware Training Course: Vmware Training Duration: 25 Days (Regular) Mode of Training: Classroom (Instructor-Led) Virtualization has redefined the way IT resources are consumed and services are delivered.

More information

Performance And Scalability In Oracle9i And SQL Server 2000

Performance And Scalability In Oracle9i And SQL Server 2000 Performance And Scalability In Oracle9i And SQL Server 2000 Presented By : Phathisile Sibanda Supervisor : John Ebden 1 Presentation Overview Project Objectives Motivation -Why performance & Scalability

More information

Deputy Secretary for Information Technology Date Issued: November 20, 2009 Date Revised: December 20, 2010. Revision History Description:

Deputy Secretary for Information Technology Date Issued: November 20, 2009 Date Revised: December 20, 2010. Revision History Description: Information Technology Policy Commonwealth of Pennsylvania Governor's Office of Administration/Office for Information Technology ITP Number: ITP-SYM008 ITP Title: Server Virtualization Policy Issued by:

More information

CUDA in the Cloud Enabling HPC Workloads in OpenStack With special thanks to Andrew Younge (Indiana Univ.) and Massimo Bernaschi (IAC-CNR)

CUDA in the Cloud Enabling HPC Workloads in OpenStack With special thanks to Andrew Younge (Indiana Univ.) and Massimo Bernaschi (IAC-CNR) CUDA in the Cloud Enabling HPC Workloads in OpenStack John Paul Walters Computer Scien5st, USC Informa5on Sciences Ins5tute jwalters@isi.edu With special thanks to Andrew Younge (Indiana Univ.) and Massimo

More information

Advances in Virtualization In Support of In-Memory Big Data Applications

Advances in Virtualization In Support of In-Memory Big Data Applications 9/29/15 HPTS 2015 1 Advances in Virtualization In Support of In-Memory Big Data Applications SCALE SIMPLIFY OPTIMIZE EVOLVE Ike Nassi Ike.nassi@tidalscale.com 9/29/15 HPTS 2015 2 What is the Problem We

More information

Mesos: A Platform for Fine- Grained Resource Sharing in Data Centers (II)

Mesos: A Platform for Fine- Grained Resource Sharing in Data Centers (II) UC BERKELEY Mesos: A Platform for Fine- Grained Resource Sharing in Data Centers (II) Anthony D. Joseph LASER Summer School September 2013 My Talks at LASER 2013 1. AMP Lab introduction 2. The Datacenter

More information

Pros and Cons of HPC Cloud Computing

Pros and Cons of HPC Cloud Computing CloudStat 211 Pros and Cons of HPC Cloud Computing Nils gentschen Felde Motivation - Idea HPC Cluster HPC Cloud Cluster Management benefits of virtual HPC Dynamical sizing / partitioning Loadbalancing

More information

Run-Time Deep Virtual Machine Introspection & Its Applications

Run-Time Deep Virtual Machine Introspection & Its Applications Run-Time Deep Virtual Machine Introspection & Its Applications Jennia Hizver Computer Science Department Stony Brook University, NY, USA Tzi-cker Chiueh Cloud Computing Center Industrial Technology Research

More information

An Implementation of Active Data Technology

An Implementation of Active Data Technology White Paper by: Mario Morfin, PhD Terri Chu, MEng Stephen Chen, PhD Robby Burko, PhD Riad Hartani, PhD An Implementation of Active Data Technology October 2015 In this paper, we build the rationale for

More information

Cloud Server. Parallels. An Introduction to Operating System Virtualization and Parallels Cloud Server. White Paper. www.parallels.

Cloud Server. Parallels. An Introduction to Operating System Virtualization and Parallels Cloud Server. White Paper. www.parallels. Parallels Cloud Server White Paper An Introduction to Operating System Virtualization and Parallels Cloud Server www.parallels.com Table of Contents Introduction... 3 Hardware Virtualization... 3 Operating

More information

VIRTUAL MACHINE LOGBOOK

VIRTUAL MACHINE LOGBOOK VIRTUAL MACHINE LOGBOOK DIPLOMA PROJECT SUMMER-FALL 2008 TASKS TO REALIZE August 6, 2008 STUDENTS: SUPERVISORS: EXPERT: ANDREA CAVALLI JULEN POFFET FRÉDÉRIC BAPST PAOLO CALAFIURA OTTAR JOHNSEN YUSHU YAO

More information

Performance Tuning of Virtual Servers TAC9872. John A. Davis Senior Consulting Engineer

Performance Tuning of Virtual Servers TAC9872. John A. Davis Senior Consulting Engineer Performance Tuning of Virtual Servers TAC9872 John A. Davis Senior Consulting Engineer Introduction Main Goals: Discuss Performance Tuning of Virtual Servers Involving ESX Server and Virtual Center Environments

More information

There are a number of factors that increase the risk of performance problems in complex computer and software systems, such as e-commerce systems.

There are a number of factors that increase the risk of performance problems in complex computer and software systems, such as e-commerce systems. ASSURING PERFORMANCE IN E-COMMERCE SYSTEMS Dr. John Murphy Abstract Performance Assurance is a methodology that, when applied during the design and development cycle, will greatly increase the chances

More information

Virtualization. P. A. Wilsey. The text highlighted in green in these slides contain external hyperlinks. 1 / 16

Virtualization. P. A. Wilsey. The text highlighted in green in these slides contain external hyperlinks. 1 / 16 1 / 16 Virtualization P. A. Wilsey The text highlighted in green in these slides contain external hyperlinks. 2 / 16 Conventional System Viewed as Layers This illustration is a common presentation of the

More information

vcenter Server 6.0: Deep Dive Nick Marshall Integration Architect 2014 VMware Inc. All rights reserved.

vcenter Server 6.0: Deep Dive Nick Marshall Integration Architect 2014 VMware Inc. All rights reserved. vcenter Server 6.0: Deep Dive Nick Marshall Integration Architect 2014 VMware Inc. All rights reserved. Who is this guy? Nick Marshall Author: Mastering vsphere Blog: NickMarshall.com.au Twitter: @NickMarshall9

More information

Windows Server 2008 R2 Hyper-V Live Migration

Windows Server 2008 R2 Hyper-V Live Migration Windows Server 2008 R2 Hyper-V Live Migration White Paper Published: August 09 This is a preliminary document and may be changed substantially prior to final commercial release of the software described

More information

Ecole des Mines de Nantes. Journée Thématique Emergente "aspects énergétiques du calcul"

Ecole des Mines de Nantes. Journée Thématique Emergente aspects énergétiques du calcul Ecole des Mines de Nantes Entropy Journée Thématique Emergente "aspects énergétiques du calcul" Fabien Hermenier, Adrien Lèbre, Jean Marc Menaud menaud@mines-nantes.fr Outline Motivation Entropy project

More information

2) Xen Hypervisor 3) UEC

2) Xen Hypervisor 3) UEC 5. Implementation Implementation of the trust model requires first preparing a test bed. It is a cloud computing environment that is required as the first step towards the implementation. Various tools

More information

Parallels Virtuozzo Containers

Parallels Virtuozzo Containers Parallels Virtuozzo Containers White Paper Greener Virtualization www.parallels.com Version 1.0 Greener Virtualization Operating system virtualization by Parallels Virtuozzo Containers from Parallels is

More information

How To Write A Monitoring System For Free

How To Write A Monitoring System For Free Zabbix : Interview of Alexei Vladishev Monitoring-fr : Hello Alexei Vladishev, can you introduce yourself to the French community please? Alexei Vladishev : I am a 36 year old engineer with a background

More information

Black-box and Gray-box Strategies for Virtual Machine Migration

Black-box and Gray-box Strategies for Virtual Machine Migration Black-box and Gray-box Strategies for Virtual Machine Migration Wood, et al (UMass), NSDI07 Context: Virtual Machine Migration 1 Introduction Want agility in server farms to reallocate resources devoted

More information

Windows Server 2008 R2 Hyper-V Live Migration

Windows Server 2008 R2 Hyper-V Live Migration Windows Server 2008 R2 Hyper-V Live Migration Table of Contents Overview of Windows Server 2008 R2 Hyper-V Features... 3 Dynamic VM storage... 3 Enhanced Processor Support... 3 Enhanced Networking Support...

More information

Microsoft SQL Server versus IBM DB2 Comparison Document (ver 1) A detailed Technical Comparison between Microsoft SQL Server and IBM DB2

Microsoft SQL Server versus IBM DB2 Comparison Document (ver 1) A detailed Technical Comparison between Microsoft SQL Server and IBM DB2 Microsoft SQL Server versus IBM DB2 Comparison Document (ver 1) A detailed Technical Comparison between Microsoft SQL Server and IBM DB2 Technical Overview about both the product offerings and their features.

More information

Virtualization and the U2 Databases

Virtualization and the U2 Databases Virtualization and the U2 Databases Brian Kupzyk Senior Technical Support Engineer for Rocket U2 Nik Kesic Lead Technical Support for Rocket U2 Opening Procedure Orange arrow allows you to manipulate the

More information

SQL Server Virtualization 101. David Klee, Group Principal and Practice Lead. SQL PASS Virtualization VC, 2014.01.08

SQL Server Virtualization 101. David Klee, Group Principal and Practice Lead. SQL PASS Virtualization VC, 2014.01.08 SQL Server Virtualization 101 David Klee, Group Principal and Practice Lead SQL PASS Virtualization VC, 2014.01.08 www.linchpinpeople.com 1 David Klee Group Principal and Practice Lead @kleegeek davidklee.net

More information

Data Center Network Minerals Tutorial

Data Center Network Minerals Tutorial 7/1/9 vmanage: Loosely Coupled Platform and Virtualization Management in Data Centers Sanjay Kumar (Intel), Vanish Talwar (HP Labs), Vibhore Kumar (IBM Research), Partha Ranganathan (HP Labs), Karsten

More information

HP Virtualization Performance Viewer

HP Virtualization Performance Viewer HP Virtualization Performance Viewer Efficiently detect and troubleshoot performance issues in virtualized environments Jean-François Muller - Principal Technical Consultant - jeff.muller@hp.com HP Business

More information

Optimizing service availability in VoIP signaling networks, by decoupling query handling in an asynchronous RPC manner

Optimizing service availability in VoIP signaling networks, by decoupling query handling in an asynchronous RPC manner Optimizing service availability in VoIP signaling networks, by decoupling query handling in an asynchronous RPC manner Voichiţa Almăşan and Iosif Ignat Technical University of Cluj-Napoca Computer Science

More information

Data Center Specific Thermal and Energy Saving Techniques

Data Center Specific Thermal and Energy Saving Techniques Data Center Specific Thermal and Energy Saving Techniques Tausif Muzaffar and Xiao Qin Department of Computer Science and Software Engineering Auburn University 1 Big Data 2 Data Centers In 2013, there

More information

High Availability of the Polarion Server

High Availability of the Polarion Server Polarion Software CONCEPT High Availability of the Polarion Server Installing Polarion in a high availability environment Europe, Middle-East, Africa: Polarion Software GmbH Hedelfinger Straße 60 70327

More information

VMWare Workstation 11 Installation MICROSOFT WINDOWS SERVER 2008 R2 STANDARD ENTERPRISE ED.

VMWare Workstation 11 Installation MICROSOFT WINDOWS SERVER 2008 R2 STANDARD ENTERPRISE ED. VMWare Workstation 11 Installation MICROSOFT WINDOWS SERVER 2008 R2 STANDARD ENTERPRISE ED. Starting Vmware Workstation Go to the start menu and start the VMware Workstation program. *If you are using

More information

OpenProdoc. Benchmarking the ECM OpenProdoc v 0.8. Managing more than 200.000 documents/hour in a SOHO installation. February 2013

OpenProdoc. Benchmarking the ECM OpenProdoc v 0.8. Managing more than 200.000 documents/hour in a SOHO installation. February 2013 OpenProdoc Benchmarking the ECM OpenProdoc v 0.8. Managing more than 200.000 documents/hour in a SOHO installation. February 2013 1 Index Introduction Objectives Description of OpenProdoc Test Criteria

More information

Ceph Distributed Storage for the Cloud An update of enterprise use-cases at BMW

Ceph Distributed Storage for the Cloud An update of enterprise use-cases at BMW Ceph Distributed Storage for the Cloud An update of enterprise use-cases at BMW Andreas Pöschl, BMW Senior Solutions Architect andreas.poeschl@bmw.de Michael Vonderbecke, BMW Solutions Architect michael.vonderbecke@bmwmc.com

More information

Big Data Management in the Clouds and HPC Systems

Big Data Management in the Clouds and HPC Systems Big Data Management in the Clouds and HPC Systems Hemera Final Evaluation Paris 17 th December 2014 Shadi Ibrahim Shadi.ibrahim@inria.fr Era of Big Data! Source: CNRS Magazine 2013 2 Era of Big Data! Source:

More information

Architecting for the next generation of Big Data Hortonworks HDP 2.0 on Red Hat Enterprise Linux 6 with OpenJDK 7

Architecting for the next generation of Big Data Hortonworks HDP 2.0 on Red Hat Enterprise Linux 6 with OpenJDK 7 Architecting for the next generation of Big Data Hortonworks HDP 2.0 on Red Hat Enterprise Linux 6 with OpenJDK 7 Yan Fisher Senior Principal Product Marketing Manager, Red Hat Rohit Bakhshi Product Manager,

More information

Xen @ Google. Iustin Pop, <iustin@google.com> Google Switzerland. Sponsored by:

Xen @ Google. Iustin Pop, <iustin@google.com> Google Switzerland. Sponsored by: Xen @ Google Iustin Pop, Google Switzerland Sponsored by: & & Introduction Talk overview Corporate infrastructure Overview Use cases Technology Open source components Internal components

More information

DELL. Virtual Desktop Infrastructure Study END-TO-END COMPUTING. Dell Enterprise Solutions Engineering

DELL. Virtual Desktop Infrastructure Study END-TO-END COMPUTING. Dell Enterprise Solutions Engineering DELL Virtual Desktop Infrastructure Study END-TO-END COMPUTING Dell Enterprise Solutions Engineering 1 THIS WHITE PAPER IS FOR INFORMATIONAL PURPOSES ONLY, AND MAY CONTAIN TYPOGRAPHICAL ERRORS AND TECHNICAL

More information